Mem Question/Kony

[dos]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

I recently bought some PQI 2-2-2-5.Had a bad module, now the replacements
run @ 2.5-4-4-8..are my replacements labled wrong?. all settings on my DFI
mobo are auto.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 8 Apr 2005 20:31:44 -0400, "doS"
<kobo65@hotREmoveThismail.com> wrote:

>I recently bought some PQI 2-2-2-5.Had a bad module, now the replacements
>run @ 2.5-4-4-8..are my replacements labled wrong?. all settings on my DFI
>mobo are auto.
>

I don't know.

If the modules are spec'd as 2,2,2,5, then that should be
programmed into their SPD (onboard prom chip). The
motherboard then reads this data... if the programmed data
is wrong, OR if the motherboard rejects/ignores it, or if
user settings override it (which it "seems" wouldn't happen
with "auto" mode if auto works properly), then you won't be
running at 2,2,2,5 timings.

I don't recall what board you have, but what I'd do is set
the memory bus synchronous to (same speed as) the FSB. If
there are any doubts that it's not doing so already, set it
that way manually. "Some" boards will set the memory bus
+33MHz higher than CPU FSB, if the CPU isn't exploiting the
fastest FSB setting the mobo chipset supports. That doesn't
mean it's better to run async memory though.

So try setting memory bus same as FSB, and manually setting
the timings to match what's on the memory sticker or specs,
the 2,2,2,5 setting. Do not boot to the OS yet, be sure to
run memtest86 and/or memtest86+ for several passes,
preferribly several hours first. Only if there are no
errors should you boot windows/other OS. If it won't do
2,2,2,5 stable at the spec'd bus speed (it was PC3200 memory
right(?) so 200MHz/DDR400 bus speed) then the memory is not
going to work and should be returned. I vaguely recall
mentioning something like this previously, that it would
probably be best to go with a different make/model of
memory, not same thing as the replacement.

This scenario is a problem in the memory industry, that
modules are being spec'd at speeds they might run stable on
a very limited number of platforms and voltages, not
necessarily "valid" specs. Most often the major brand
names would sell modules of "same" potential performance
with lower specs for the part (higher numbered, slower
timings). IOW, PDI can't buy same or fewer numbers of
2,2,2,5 chips at lower price than anybody else, the market
is too competitive for that to happen for a smaller
manufacturer/relabeler.

 

[dos]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

So,simply the board may not run these settings at default,it may take some
tweaking to get the 2-2-2-5 if desired.The system is stable at a very minor
overclock, I have the FSB @ 210, it's an AMD 64 3400+,DFI LANPARTY UT 250GB
Nforce 3 board. The Ram is PQI 3200 2-2-2-5 .The returned memory had errors
in test 5,one module, but as you suggested, I returned both sticks for a new
set. I seem to think the good stick of returned ram defaulted at
2-3-2-5.Thanks for your help, as I said, the initial memory was returned
because one module had errors all through out test 5.



"kony" <spam@spam.com> wrote in message
news:bf9e511hvqcif9usslidlua71duj45vkrf@4ax.com...
> On Fri, 8 Apr 2005 20:31:44 -0400, "doS"
> <kobo65@hotREmoveThismail.com> wrote:
>
> >I recently bought some PQI 2-2-2-5.Had a bad module, now the replacements
> >run @ 2.5-4-4-8..are my replacements labled wrong?. all settings on my
DFI
> >mobo are auto.
> >
>
> I don't know.
>
> If the modules are spec'd as 2,2,2,5, then that should be
> programmed into their SPD (onboard prom chip). The
> motherboard then reads this data... if the programmed data
> is wrong, OR if the motherboard rejects/ignores it, or if
> user settings override it (which it "seems" wouldn't happen
> with "auto" mode if auto works properly), then you won't be
> running at 2,2,2,5 timings.
>
> I don't recall what board you have, but what I'd do is set
> the memory bus synchronous to (same speed as) the FSB. If
> there are any doubts that it's not doing so already, set it
> that way manually. "Some" boards will set the memory bus
> +33MHz higher than CPU FSB, if the CPU isn't exploiting the
> fastest FSB setting the mobo chipset supports. That doesn't
> mean it's better to run async memory though.
>
> So try setting memory bus same as FSB, and manually setting
> the timings to match what's on the memory sticker or specs,
> the 2,2,2,5 setting. Do not boot to the OS yet, be sure to
> run memtest86 and/or memtest86+ for several passes,
> preferribly several hours first. Only if there are no
> errors should you boot windows/other OS. If it won't do
> 2,2,2,5 stable at the spec'd bus speed (it was PC3200 memory
> right(?) so 200MHz/DDR400 bus speed) then the memory is not
> going to work and should be returned. I vaguely recall
> mentioning something like this previously, that it would
> probably be best to go with a different make/model of
> memory, not same thing as the replacement.
>
> This scenario is a problem in the memory industry, that
> modules are being spec'd at speeds they might run stable on
> a very limited number of platforms and voltages, not
> necessarily "valid" specs. Most often the major brand
> names would sell modules of "same" potential performance
> with lower specs for the part (higher numbered, slower
> timings). IOW, PDI can't buy same or fewer numbers of
> 2,2,2,5 chips at lower price than anybody else, the market
> is too competitive for that to happen for a smaller
> manufacturer/relabeler.

 

[dos]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

General
Memory type DDR-SDRAM
Manufacturer (ID) (0000000000000000)
Size 512 MBytes
Max bandwidth PC3200 (200 MHz)
Part number
Manufacturing date Week 00/Year 00

Attributes
Number of banks 2
Data width 64 bits
Correction None
Registered no
Buffered no

Timings table
Frequency (MHz) 200 200
CAS# 2.5 3.0
RAS# to CAS# delay 4 4
RAS# Precharge 4 4
TRAS# 8 8
"kony" <spam@spam.com> wrote in message
news:bf9e511hvqcif9usslidlua71duj45vkrf@4ax.com...
> On Fri, 8 Apr 2005 20:31:44 -0400, "doS"
> <kobo65@hotREmoveThismail.com> wrote:
>
> >I recently bought some PQI 2-2-2-5.Had a bad module, now the replacements
> >run @ 2.5-4-4-8..are my replacements labled wrong?. all settings on my
DFI
> >mobo are auto.
> >
>
> I don't know.
>
> If the modules are spec'd as 2,2,2,5, then that should be
> programmed into their SPD (onboard prom chip). The
> motherboard then reads this data... if the programmed data
> is wrong, OR if the motherboard rejects/ignores it, or if
> user settings override it (which it "seems" wouldn't happen
> with "auto" mode if auto works properly), then you won't be
> running at 2,2,2,5 timings.
>
> I don't recall what board you have, but what I'd do is set
> the memory bus synchronous to (same speed as) the FSB. If
> there are any doubts that it's not doing so already, set it
> that way manually. "Some" boards will set the memory bus
> +33MHz higher than CPU FSB, if the CPU isn't exploiting the
> fastest FSB setting the mobo chipset supports. That doesn't
> mean it's better to run async memory though.
>
> So try setting memory bus same as FSB, and manually setting
> the timings to match what's on the memory sticker or specs,
> the 2,2,2,5 setting. Do not boot to the OS yet, be sure to
> run memtest86 and/or memtest86+ for several passes,
> preferribly several hours first. Only if there are no
> errors should you boot windows/other OS. If it won't do
> 2,2,2,5 stable at the spec'd bus speed (it was PC3200 memory
> right(?) so 200MHz/DDR400 bus speed) then the memory is not
> going to work and should be returned. I vaguely recall
> mentioning something like this previously, that it would
> probably be best to go with a different make/model of
> memory, not same thing as the replacement.
>
> This scenario is a problem in the memory industry, that
> modules are being spec'd at speeds they might run stable on
> a very limited number of platforms and voltages, not
> necessarily "valid" specs. Most often the major brand
> names would sell modules of "same" potential performance
> with lower specs for the part (higher numbered, slower
> timings). IOW, PDI can't buy same or fewer numbers of
> 2,2,2,5 chips at lower price than anybody else, the market
> is too competitive for that to happen for a smaller
> manufacturer/relabeler.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 8 Apr 2005 21:00:54 -0400, "doS"
<kobo65@hotREmoveThismail.com> wrote:

>So,simply the board may not run these settings at default,it may take some
>tweaking to get the 2-2-2-5 if desired.The system is stable at a very minor
>overclock, I have the FSB @ 210, it's an AMD 64 3400+,DFI LANPARTY UT 250GB
>Nforce 3 board. The Ram is PQI 3200 2-2-2-5 .The returned memory had errors
>in test 5,one module, but as you suggested, I returned both sticks for a new
>set. I seem to think the good stick of returned ram defaulted at
>2-3-2-5.Thanks for your help, as I said, the initial memory was returned
>because one module had errors all through out test 5.
>

TO determine if the memory is working properly you must
return the system to non-o'c, spec'd speed. With memory at
correct bus speed, and only then, the memory should pass all
memory tests. It may well be that increasing the FSB to
oveclock will cause the bios to automatically choose higher,
slower timings. You could now still manually set 2,2,2,5 at
your present o'c bus speeds- at your own risk since it's
certainly not guaranteed to work.

You might be better off with the spec'd FSB speed too, since
it will (if the memory works properly) be similar or
sometimes faster memory performance... something crude
memory benchmarks which merely show MB/s, may not reflect
accurately. On the other hand, if you have no other way to
increase the CPU clock speed- that the CPU multiplier is
locked, then that might be a justification to continue with
the higher bus speed.

 

[dos]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

I reset the FSB to 200, which is default,I am trying to determine in my bios
exactly where to adjust the 2-2-2-5 settings.

http://webpages.charter.net/dos/DSC00950e.jpg



"kony" <spam@spam.com> wrote in message
news:8dbe51lug3uscqgv184o1kt8l022ca7n32@4ax.com...
> On Fri, 8 Apr 2005 21:00:54 -0400, "doS"
> <kobo65@hotREmoveThismail.com> wrote:
>
> >So,simply the board may not run these settings at default,it may take
some
> >tweaking to get the 2-2-2-5 if desired.The system is stable at a very
minor
> >overclock, I have the FSB @ 210, it's an AMD 64 3400+,DFI LANPARTY UT
250GB
> >Nforce 3 board. The Ram is PQI 3200 2-2-2-5 .The returned memory had
errors
> >in test 5,one module, but as you suggested, I returned both sticks for a
new
> >set. I seem to think the good stick of returned ram defaulted at
> >2-3-2-5.Thanks for your help, as I said, the initial memory was returned
> >because one module had errors all through out test 5.
> >
>
> TO determine if the memory is working properly you must
> return the system to non-o'c, spec'd speed. With memory at
> correct bus speed, and only then, the memory should pass all
> memory tests. It may well be that increasing the FSB to
> oveclock will cause the bios to automatically choose higher,
> slower timings. You could now still manually set 2,2,2,5 at
> your present o'c bus speeds- at your own risk since it's
> certainly not guaranteed to work.
>
> You might be better off with the spec'd FSB speed too, since
> it will (if the memory works properly) be similar or
> sometimes faster memory performance... something crude
> memory benchmarks which merely show MB/s, may not reflect
> accurately. On the other hand, if you have no other way to
> increase the CPU clock speed- that the CPU multiplier is
> locked, then that might be a justification to continue with
> the higher bus speed.
>
>

 

[dos]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

I dont see RAS# Precharge , is that the same as Row Precharge on my screen?

"kony" <spam@spam.com> wrote in message
news:8dbe51lug3uscqgv184o1kt8l022ca7n32@4ax.com...
> On Fri, 8 Apr 2005 21:00:54 -0400, "doS"
> <kobo65@hotREmoveThismail.com> wrote:
>
> >So,simply the board may not run these settings at default,it may take
some
> >tweaking to get the 2-2-2-5 if desired.The system is stable at a very
minor
> >overclock, I have the FSB @ 210, it's an AMD 64 3400+,DFI LANPARTY UT
250GB
> >Nforce 3 board. The Ram is PQI 3200 2-2-2-5 .The returned memory had
errors
> >in test 5,one module, but as you suggested, I returned both sticks for a
new
> >set. I seem to think the good stick of returned ram defaulted at
> >2-3-2-5.Thanks for your help, as I said, the initial memory was returned
> >because one module had errors all through out test 5.
> >
>
> TO determine if the memory is working properly you must
> return the system to non-o'c, spec'd speed. With memory at
> correct bus speed, and only then, the memory should pass all
> memory tests. It may well be that increasing the FSB to
> oveclock will cause the bios to automatically choose higher,
> slower timings. You could now still manually set 2,2,2,5 at
> your present o'c bus speeds- at your own risk since it's
> certainly not guaranteed to work.
>
> You might be better off with the spec'd FSB speed too, since
> it will (if the memory works properly) be similar or
> sometimes faster memory performance... something crude
> memory benchmarks which merely show MB/s, may not reflect
> accurately. On the other hand, if you have no other way to
> increase the CPU clock speed- that the CPU multiplier is
> locked, then that might be a justification to continue with
> the higher bus speed.
>
>

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

In article <40G5e.125$2A6.69@fe07.lga>, doS says...
> General
> Memory type DDR-SDRAM
> Manufacturer (ID) (0000000000000000)
> Size 512 MBytes
> Max bandwidth PC3200 (200 MHz)
> Part number
> Manufacturing date Week 00/Year 00
>
> Attributes
> Number of banks 2
> Data width 64 bits
> Correction None
> Registered no
> Buffered no
>
> Timings table
> Frequency (MHz) 200 200
> CAS# 2.5 3.0
> RAS# to CAS# delay 4 4
> RAS# Precharge 4 4
> TRAS# 8 8


Take it back. If you bought CAS 2 stuff originally, then the above is
not a direct replacement. Looking at the default timings of the
original ad replacement, the replacements are way slower and I would
suspect priced far cheaper too.

--
Conor

Windows & Outlook/OE in particular, shipped with settings making them
as open to entry as a starlet in a porno. Steve B

 

[dos]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Thanks, but setting those values, the 2-2-2-5 in the bios works fine, ran
memetest for 11 hrs, no errors.

"Conor Turton" <conor@conorturton.com> wrote in message
news:MPG.1cc1b85bfc7a7b3b989771@news.individual.net...
> In article <40G5e.125$2A6.69@fe07.lga>, doS says...
> > General
> > Memory type DDR-SDRAM
> > Manufacturer (ID) (0000000000000000)
> > Size 512 MBytes
> > Max bandwidth PC3200 (200 MHz)
> > Part number
> > Manufacturing date Week 00/Year 00
> >
> > Attributes
> > Number of banks 2
> > Data width 64 bits
> > Correction None
> > Registered no
> > Buffered no
> >
> > Timings table
> > Frequency (MHz) 200 200
> > CAS# 2.5 3.0
> > RAS# to CAS# delay 4 4
> > RAS# Precharge 4 4
> > TRAS# 8 8
>
>
> Take it back. If you bought CAS 2 stuff originally, then the above is
> not a direct replacement. Looking at the default timings of the
> original ad replacement, the replacements are way slower and I would
> suspect priced far cheaper too.
>
> --
> Conor
>
> Windows & Outlook/OE in particular, shipped with settings making them
> as open to entry as a starlet in a porno. Steve B